Bid the Market
Hedge Systems, Inc
To Run the software, input parameter into the following form and click the summit button
In
the past several years, the equity achieves great return and huge volatility.
Portfolio managers and individual investors want to bid the market and reduce
the volatility.
In
the past year, the growth concentrated in the new technology, such as Internet,
Communication. The majority of the stock is in the negative area. The market
would be a bear market without the new Technology section.
The
traditional method to select stock and track the index was very poorly. Most of
the growth fund did not bid the index.
In
this paper, we outline a methodology to select stock.
1.
The
market
In
the following graph, we assume invest $100 at five indexes, Dow Jones Industry
Average(^DJI), NAS/NMS COMPSITE(^IXIC), S&P 100 INDEX(^OEX), S&P 500
INDEX(^SPX) and RUSSELL 2000(^RUT). The result at 3/26/99 is Dow Jones $124.757
, NAS/NMS COMPSITE is 152.1357. S&P 100 INDEX is 131.0852. S&P 100 INDEX is 138.4705 and RUSSELL 2000 is 91.33106. At same period,
At Same Period, WCOM increase 285.32 percent. Microsoft
(MSFT) increase 258.50 percent and American on Line (AOL) increase 1048.76
percent.
In the past year, the broad market is far behind. The
growth is concentrated a handsome of new technologies. Though there is a
probability that the broad market may drag down the growth of the high flied
securities, there are higher chances the fly will continue due to worldwide
recovery.
Regression model is widely used in the process to select
stocks. However, it seems less successful in recent years. We intend to improve
the regression model to fit the current market conditions. First we select a
set of indexes that we think better reflect the current market conditions. Then
we create multiple step regression models to adapt the markets. As the result,
we estimate the parameter a, which represents the
daily excess return over the market for specified securities. To compare the
return, we define the risk-adjusted return as ration of the excess return over
the specific volatility.
2.
Select Indexes
To reflect the behavior of the markets, we should select
different set of indexes. We select the following set of indexes.
|
Ticker |
Description |
|
^BIX |
S&P BANKINDEX |
|
^CWX |
CBOE COMPSFTWR |
|
^DJI |
DJ INDUAVERAGE |
|
^DRG |
PHARMCEUTICLNDX |
|
^HCX |
S&P HLTHCARE |
|
^INX2 |
INTERNET INDEX |
|
^IUX |
S&P INSURANCE |
|
^IXIC |
NAS/NMS COMPSITE |
|
^MSH |
MSTAN HITECH |
|
^OEX |
S&P 100INDEX |
|
^OIX |
OIL INDEX |
|
^RLX |
S&P RETAILNDX |
|
^RUT |
RUSSELL 2000 |
|
^SPX |
S&P 500
INDEX |
|
^VIX |
MKTVOLTLTY NDX |
|
^XAL |
AIRLINE INDEX |
|
^XBD |
SEC BROKER DEALR |
|
^XCI |
COMPUTER TECH |
The
internet index and CBOE Computer Software Index reflect the most active sector
on the market. The Market Volatility Index gives the short posture of the
markets.
3.
The
model
![]()
We assume the price of
equity follows a lognormal model. In other works the return follows the normal
model. The current return is a liner combination of past m steps of the index
returns. Mathematically, we can write the equity return as
Where
X(k,i-l) is the index k return at time i-l.
![]()
Suppose we observed a
sequence of returns of the index and the equity. The coefficient a and b can be identify by making
the noise as small as possible. Mathematically, we will solve the following
problem:
The
above problem is called adapted parameter identification problem. There are
several method can solve above equations.
The
following graph shows the project prices and actual market prices. The length
of time sequence is 300 day. The delay period is 2.

4.
Adjusted
Excess Return
Parameter
a represents the daily excess return of the
equity with respect to the market. The specific volatility s represents the volatility which can not
explained by the market. Since the risk
level is different, the return of different equity can not compare each other.
Therefore we define the risk adjusted excess adjusted return as the ration of a and s. Mathematically,
Risk
Adjusted Excess Return m
![]()
The
following table shows the excess return, specific volatility and risk adjusted
excess returns.

In
above table, The AOL and MSFT has the highest Risk Adjusted Return. The excess return of YHOO is higher than
Microsoft. However the specific volatility is much higher than Microsoft.
Consequently, the risk-adjusted return are lower than Microsoft.
5.
Value
at Risk
The
above model can be used as value at risk. The sum of b is the sensitivity of the security with
respect to the indexes.
To
evaluate the value at risk of a equity, we can decompose the risk into the
market risk and the specific risk. The market risk can further divided into the
risk due to different indexes.
6.
Assets
allocations
The
risk adjusted excess return can be the basis of assets allocation. We suggest
to hold a portfolio with the ratio proportion to the risk adjusted excess
return.
Suppose
we have $100, we will invest our money into the following portfolio as 3/26/99.
|
AOL |
24.16878 |
|
MSFT |
18.4713 |
|
YHOO |
8.698968 |
|
BA |
3.235848 |
|
IBM |
7.512935 |
|
ATHM |
6.521202 |
|
DELL |
12.28776 |
|
WCOM |
19.1032 |